Talk to an Expert
Talk to an Expert ✆ +91 945 945 6700
Stock Audit · 6 min read · Aug 19, 2026

What Does a Warehouse Auditor Do? What an Independent Count Tests, Not Your WMS

CA Sundram Gupta

What Does a Warehouse Auditor Do? What an Independent Count Tests, Not Your WMS - Featured Image
In this guide

    What an Independent Count Tests

    An independent count tests four assertions about the goods: that they exist, that the business owns them, that they are in a condition to be sold, and that they are carried at a value the market supports. Existence is tested by counting. Ownership is tested against purchase records, storage agreements and the identity of the principal where the goods belong to somebody else. Condition is assessed on the floor. Valuation is tested against cost and against what the goods will realise. What a warehouse management system cannot do is self-attest to any of these. The system holds what it was told, and every one of the four assertions can be false while the system remains internally consistent: goods recorded and never received, goods belonging to a principal recorded as owned, damaged stock carried at full condition, and obsolete lines at full cost. A system report is a claim by the business about itself, which is the definition of what independent evidence is not.

    What Your WMS Already Tells You

    A warehouse management system is a considerable source of information and it is worth being precise about which parts of it are reliable. System stock by location is its core output, and it is authoritative about what the system was told: every receipt booked, every put-away confirmed, every pick recorded and every adjustment posted. Where transaction discipline is good, that is a close description of the building. Movement history is the system's real strength and the part no count can reproduce. It holds when each item arrived, where it was placed, when it was picked and against which order, which is exactly what allows a difference found at a count to be investigated rather than merely reported. An auditor who ignores the movement history is discarding the most useful evidence available. Where the system is authoritative and where it is not is the distinction that matters. It is authoritative about transactions and sequences. It is not authoritative about the physical world, because it has no sensor: it knows what was entered, and every unrecorded movement, mis-scan, unbooked receipt and unlogged damage is invisible to it while leaving the recorded balance perfectly self-consistent.

    What Only a Physical Count Establishes

    Three things cannot be established from any record and require somebody standing in the building. Stock that exists against stock recorded is the first and most obvious. Only observation confirms that the goods the system carries are physically present, and the whole logic of verification is that a record cannot corroborate itself. A system reporting a quantity is repeating what it was told. Condition and saleability is the second, and it is entirely absent from most systems. A warehouse management system holds quantities and locations; it does not know that a pallet has been crushed, that a carton has been water-damaged, or that stock has passed the date after which the trade will accept it. Goods in that condition are counted as present and are not worth what the records say. Ownership of what is on the floor is the third and it is the one that matters most to a lender. A facility routinely holds goods belonging to customers, consignors and principals alongside its own, and physical presence establishes nothing about title, so a count producing a total without establishing whose each part is has produced a number nobody can advance against.

    Independence and Why It Matters to a Lender

    The value of a count to an outside reader depends on who performed it, which is a point about evidence rather than about competence. Self-counting against independent counting is the distinction. A count performed by the team responsible for the stock may be entirely careful and it remains a statement by a party about its own position, made by people whose performance is measured by the result. That is precisely the situation external evidence exists to address, and no amount of thoroughness converts it into independent evidence. Who signs and what that signature means is the second element. A report signed by a Chartered Accountant in practice carries the professional standards, the continuing obligations and the disciplinary framework attaching to that membership, which is what allows a lender to rely on a document prepared by somebody it has never met. UDIN completes the chain on an assurance report. The Unique Document Identification Number generated by the signing member allows the recipient to confirm through the Institute's portal that the document was issued by the member named on it, which closes the one gap a signature alone leaves open.

    Evidence the Auditor Produces

    An independent count produces three artefacts, and the first is the one that makes it evidence rather than activity. Reconciliation to book stock ties what was physically found back to the balance in the records as at the same moment, with each difference identified individually rather than netted. A count reported as a quantity, without that reconciliation, tells a reader what was on the floor and nothing about whether the accounts are right. The exception listing by class is the second, separating differences into the categories that carry different consequences: quantity differences, goods present but belonging to others, goods in a condition that affects value, and goods whose ownership could not be established. Grouping them together forces the analysis to be repeated by whoever reads it. What goes into the credit file is the third, and it is a distillation rather than the whole report: the verified value, the coverage achieved, the observations raised, and whether the position agreed with what the borrower had been reporting. That last comparison is the one carried forward and read again at the next review.

    Commissioning a Count

    The site list is the brief, and it needs more than addresses. For each location: what is held there, an approximate line or pallet count, whether the stock is yours or held for others, who controls access, and what system the site runs on. Those six items produce a scope that survives contact with the work. A request to count the warehouses, without them, is an invitation to guess. Time the count around the operation rather than the finance calendar where the two conflict, because a count attempted during peak dispatch either fails or costs more than it should. Identify the quiet window at each site and work back from it. Where a reporting date is fixed and the operation cannot pause, agree the boundary procedure in advance instead of hoping. Have four things ready: a system extract as at the cut-off, the goods movement documents around it, the storage agreements for anything held for others, and confirmed access. Warehouse stock audit work is quoted from that information and rarely varies afterwards.

    Related reading

    Share this guide: Link copied!

    What can an independent count detect that a WMS cannot?

    Anything the system was told incorrectly: stock booked in but never received, damages not recorded, put-away errors, and substitution. The WMS faithfully reports its inputs, which is exactly why it cannot validate them.

    Does a good WMS reduce the need for stock audits?

    It reduces the size of variances but not the need to verify them independently. Lenders and auditors rely on external evidence, and a system record, however sophisticated the system, remains a management representation rather than independent confirmation.

    What does the auditor test about the warehouse process?

    Goods-inward checking, put-away discipline, segregation of damaged and quarantined stock, cycle-count practice, and how variances are approved and posted to the ledger. Process weaknesses are what predict where the next variance will appear, which is why they are tested alongside the count.

    How does an auditor sample a large warehouse?

    By stratifying on value and movement, testing high-value and fast-moving lines heavily, and sampling the long tail. Random sampling across all SKUs treats a pallet of packaging like a pallet of finished goods.

    What is the difference between an internal cycle count and an audit?

    Independence. Cycle counts are performed by the team responsible for the stock; an audit is performed by someone who is not. The procedures may look similar but the assurance they provide is not the same.